Master’s Thesis Exposé

The Gateway Drug for Artificial Intelligence: How can bots be authentic?

Keywords: Intelligent Messaging Agents, Social Intelligence, Brand Authenticity

Technology gains more and more importance in people’s every-day life according to an US American survey from 2014 (Pew Research Center, 2014). The presence of social media, messaging and digital substitutes for human interaction will change the way humans are behaving in social and cultural manner (Lohr, 2015). Therefore, the domain of digital behavior will become a very attractive field of research for anthropologists, economists and technologists alike. Digging deeper into this research area of technology and human interaction by having a closer look on behavioral and business-related facets shows areas to narrow and to reveal the most emerging research aspects.

Taking Gartner’s Hype Cycle of Emerging Technologies into consideration, a momentum towards digital business and digital marketing, can be identified among the emerging aspects (Gartner Inc., 2015). Enlarging this perspective by business insights from a research done by Pew Research Center, mobile messaging and personalized communication are the ways of electronic communication that fit the needs of younger generations (Duggan, 2015). However, after a trend in the mid 90s led to the replacement of desktop OS through browsers and websites, another paradigm shift from browsers to apps has been seen in the last years (Sheth, 2015). Yet, another behavioral change is on the horizon. Beerud Sheth boldly describes this in his TechCrunch article, “just as websites replaced client applications then, messaging bots will replace mobile apps now” (Sheth, 2015). Therefore, it appears that messaging bots will become the new mean of communication for brands and consumers alike. This paradigm change will impact how brands communicate and interact with their peers.

Therefore, this Master’s Thesis will first elaborate the paradigm change towards bot-driven messaging. Secondly, characteristics of authenticity have to be identified out of literature. Different target groups are tested on hypotheses derived from the literature review on authenticity. Finally, the core motivation of this thesis is to come up with a set of characteristics enabling the development of socially intelligent messaging agents.

Theoretical Background

Contemporary marketing occurs in a complex world indelibly shaped by cultural and social forces of its manifold stakeholders (Yücel & Dagelen, 2010). Technological innovation, emergent cultural ideas, and shifts in economic and political structures mean that marketers and marketing scholars have an infinite array of new ways of interaction and new concepts for brand communication to explore. As marketing activities engage with this complex world, cultural orientation and thought-through diffusion of innovation are a crucial part in the value generation of organizations. In order to be successful, organizations need to master the infinite innovation cycle, adapt to consumers’ behavior and changes in culture (Fenn & Raskino, 2008).

The cultural and behavioral change is best showcased by the rise of digital natives and their distinction to the digital immigrant generations. The Google and millennial generations show different patters for communication and interaction with brands than earlier generations. Ever since, digital natives are influenced by the omnipresence of the Internet, multitasking and round-the-clock information accessibility. The later is fostered by the bright usage of the Internet (Prensky, 2009).

In their most recent national survey with 1907 telephone interviewees, Pew Research Center found that 36% of smartphone owners report using messaging services such as WhatsApp, Kik or iMessage to interact with and communicate to brands (Duggan, 2015). Recent announcements and news articles underpin the importance and rise of mobile and therefore indicate the importance for research in this area. In November this year for example, Chinese online market platform Alibaba Group Holding Ltd’s announced a share of 68 percent for goods ordered via mobile devices during its Singles’ Day shopping festival which showed a total value of goods transacted of $14.32 billion (Carsten, 2015). The four-year-old messaging service WeChat, a service of the Chinas largest Internet company Tencent, counts nearly 500 million monthly active users and generated about $420 million in mobile games revenue (Lohr, 2015). Other examples for the rise of mobile messaging services as facilitator of mobile brand interaction is WhatsApp, the mobile messaging service purchased by Facebook for $22 billion in late 2014. David Wehner, Facebook’s chief financial officer, revealed that the social network intents to explore business-to-consumer (B2C) interaction in WhatsApp (Haro, 2015). What these apps have in common is the outlook of integrating intelligent agents for the facilitation of communication. Therefore, it can be mentioned that for example Facebook is working on an intelligent agent called ‘M’ for its own Facebook messenger service (Olanoff & Constine, 2015). Messaging ought to be the new platform according to the above mentioned examples. Even more, also venture capitalists like Benedict Evans from Andreessen Horowitz see a change towards mobile messaging: “Old: all software expands until it includes messaging. New: all messaging expands until it includes software” (Evans, 2015). So if “messaging is the new platform”, then “bots are the new apps” according to Beerud Sheth (2015) in his article ‘Forget Apps! Message Bots Are The Future of Customer Engagement!’. However, the rise of intelligent agents is not solely noticeable for mobile messaging. For example, companies like the New York Times use intelligent agents within their messaging and collaboration environment ‘Slack’. The bot called ‘Blossom’ helps to identify articles with potential of driving virality using comprehensive data mining and prediction algorithms as well as messaging within the Slack environment (Wang, 2015). Moreover, Slack, a messaging and collaboration platform, allows organizations to use bot integrations for enterprise resource planning, customer relationship management, human resource management and other enterprise processes using simple messaging (Sheth, 2015). Therefore, messaging bots are defined as computer programs that read and write messages, just like human users. However, intelligent messaging agents can automate workflows and provide advanced services (Oskouei, Varzeghani, & Samadyar, 2014).

As described above, humans interact with intelligent agents during work and private life. Unnoticed by the user, the software agent mimics human traits (Ferrara, Varol, Davis, Menczer, & Flammini, 2015). Since the creation of intelligent systems like chatbots, code designed to hold a conversation with a human, the Turing test tries to differentiate human behavior from the behavior of computer algorithms (Turing, 1950). With the rise of a messaging ecosystem where messaging is the new platform and bots inhabit the place of apps, intelligent agents “raise the bar of the challenge, as they introduce new dimensions to emulate in addition to content, including the social network, temporal activity, diffusion patterns and sentiment expression” as described in the ‘Rise of Social Bots’ (Ferrara, Varol, Davis, Menczer, & Flammini, 2015, p. 1). The so called social bots are already well-known for their behavior out of several studies on Twitter, where social bots are detected frequently (Boshmaf, Muslukhov, Beznosov, & Ripeanu, 2013; Ferrara, Varol, Davis, Menczer, & Flammini, 2015; Lee, Eoff, & Caverlee, 2011). “A social bot is a computer algorithm that automatically produces content and interacts with humans” on a network “trying to emulate and possibly alter their behavior” (Ferrara, Varol, Davis, Menczer, & Flammini, 2015, p. 1). This definition is also applicable to the business cases mentioned above, when a messaging bot helps with orders of food, engages with customers or facilitates certain business processes. Although, coming back to the purpose of the test envisioned by Alan Turing, it is to determine how human-like computer systems behave. This includes the social capabilities of the algorithm. As Persson, Laaksolahti and Lönnqvist put in their first sentences of ‘Understanding Socially Intelligent Agents — A Multilayered Phenomenon’: “The ultimate purpose with socially intelligent agent (SIA) technology is not to simulate social intelligence per se, but to let an agent give an impression of social intelligence” (Persson, Laaksolahti, & Lönnqvist, 2001, p. 349).

Going through literature Karl Albrecht defines five key aspects of the social intelligence of human beings, which computer scientist seek to implement into algorithms up to a certain point. Leigh Rivenbark (2006) sums up Albrechts’ findings by describing the key principles of Situational Awareness, Presence, Authenticity, Clarity and Empathy. Going into more detail, situational awareness is described in the ability of understanding the situation and its circumstances creating a “social radar” to detect what is going on around a situation (Rivenbark, 2006). Studies towards situational and context awareness of intelligent agents are performed by various teams (Lei, 2005; Kornienko, Kornienko, & Levi, 2005). “Context awareness allows applications to adapt themselves to their computing environment in order to better suit the needs of the user” as defined in ‘Context Awareness: a Practitioner’s Perspective’ (Lei, 2005, p. 1). Hui Lei (2005, p. 1) further elaborates that “it (contextual awareness) promises to reduce the demand for human attention, which arguably is the most limited resource in an environment saturated with computing and communication capability”. The topic of contextual awareness also finds application in general artificial intelligence research like swarm robotics (Kornienko, Kornienko, & Levi, 2005). However, a lack in literature and research regarding the social intelligence of intelligent agents can be identified going along the line with social intelligence capabilities of presence, authenticity, clarity and empathy.

Bridging the gap from computer science to consumer behavior, honesty and authenticity are gaining in importance for consumers around the world (Cohn & Wolf, 2014). Phrased by Pine and Gilmore (2008, p. 19) “authenticity is becoming the new consumer sensibility”. This makes authenticity an important topic in marketing literature as well (Leigh, Peters, & Shelton, 2006). Authenticity has been researched in the area of tourism and learning, where several frameworks from different viewpoints have been developed (Gulikers, Bastiaens, & Kirschner, 2004; Kolar & Zabkar, 2010). However, a research gap is still present for areas such as software development and in particular the development of autonomous algorithms interacting with humans. In the early stages of intelligent agents and their rising application for business use, Johnson and Noorman (2014, p. 25) claim that “responsibility issues should be addressed when artificial agent technologies are in the early stages of development”. The responsibility for this paper is to support developers and users of intelligent agents to make them more human like by becoming authentic. A goal, which already exists since the advent of artificial intelligence (Turing, 1950).

Are U Real?

Objectives and Research Question

In terms of emergent technological innovation, this Master’s Thesis will focus on messaging as a key trend in digital marketing for the years to come (Gartner Inc., 2015). More precisely, authentic messaging behavior of intelligent agents. Messaging services fosters personalization and instant interaction between business and customers through a combination of human problem-solving and computed intelligent assistance. So called concierge services, a predecessor for artificial intelligence in messaging, are a manifestation of the readiness of this services in the real world. These services already take advantage of the possibilities of intelligent agents. Respectively, a 2015 Gartner report mentions, that “the future will belong to the companies that can create the most effective autonomous and smart software solutions” (Lohr, 2015).

This Master’s Thesis will first elaborate the paradigm change towards bot-driven messaging. Therefore, it can be necessary to determine the status quo of the adaption of mobile messaging within marketing. Moreover, the advancement of intelligent agents in behaving and acting socially intelligent have to be elaborated. Special attention has to be given to the integration of situation and context-awareness of intelligent agents since these capabilities are already discussed in literature.

Challenging the gap of missing social intelligence of bots and respectively the lack of research on how intelligent agents can become authentic, this master thesis will first elaborate how authenticity is defined by already existing literature. Out of this literature review, the aspects influencing authenticity for intelligent agents from the viewpoint of brands and developers shall be tested. The outcome of the qualitative research is to create a framework for socially intelligent agents that fit the expectations of brand managers and software developers alike which helps to prevent the creation of bad bots.

The research question deriving out of the problem background and objectives of the Master’s Thesis can be framed as follows:

What are the characteristics of an authentic intelligent messaging agent?

Methodology

For the first part, extensive literature review has to be carried out. On one side, harvesting current literature regarding intelligent agents research to determine the status quo of current developments working along the concept of social intelligence. On the other side, it is necessary to establish an understanding of the characteristics of authenticity. The literature review makes up the basis to hypothesize on characteristics of authentic intelligent messaging agents.

Respectively, two different groups are tested through qualitative interviews on the derived hypotheses. The target groups are developers and product managers of intelligent agent software as well as brand managers of brands which integrate intelligent messaging agents in their communication with customers. The aim is to work together with businesses offering concierge services like GoButler, Magic or James Bitte to get insights into the development of intelligent agents (bots). Two out of four contacted developer teams already stated interest in this research.

The sample will compromise of up to ten interviews per group. The interviews will deal with the aspects identified in the literature review indicated as important characteristics of authenticity. The outcome of the interviews will then be analyzed according to Mayrings qualitative content analysis (Mayring, 2014). The results in terms of characteristics for each of the groups are tested against the hypothesis. In a further step, the thesis aims to identify correlations as well as gaps within the answers of both groups as well as the hypotheses. The aim is to answer the research question and to come up with a holistic set of characteristics for authentic intelligent messaging agents.

The chosen methodology is defined as most appropriate, since it is estimated that a quantitative survey would not produce unbiased results. The questions and provided answers in a quantitative survey would imply a certain importance to characteristics and thereby influence the result. Further, the participants of each group are scattered all over the globe. This circumstance and the scarce number of business applying intelligent messaging agents prevents a research design making use of focus groups. However, this methodology would allow the participants of the groups to interact and enrich results.

Conclusions, Limitations and Further Research

The expectation for the survey is to get divergent results for both survey groups as well as deviations from the characteristic of authenticity form the literature review. The purpose of the thesis is to find correlations in the survey answers and gain a set of characteristics for the development of authentic intelligent agents. The inclusion of characteristics of authenticity will enable intelligent agents to be socially intelligent to a certain point and increase the acceptance and integration into other services than messaging. Since authenticity has an influence on marketing activities and the perception of customers, the research can support intelligent agent integration into marketing automation.

However, due to limited timeframe and a still small group of business using intelligent agents, the thesis has to cope with the disadvantage of a small sample. The lack of available literature and research in the area of social intelligence of bots combined with marketing approaches like authenticity is another limitation of the thesis. Therefore, the thesis has to brake ground in this field.

Further research should include the creation of dialogues for authentic messaging, after characteristics of authenticity for intelligent agents have been discovered. Moreover, further parts of the social intelligence concept can be evaluated and integrated into a framework for the development of social intelligent bots.

References

Boshmaf, Y., Muslukhov, I., Beznosov, K., & Ripeanu, M. (2013). Design and analysis of a social botnet. Computer Networks, 57(2), 556–578.

Carsten, P. (2015, 11 12). Alibaba’s Singles’ Day sales surge 60 percent to $14.3 billion. Retrieved 11 18, 2015, from Thomson Reuters: http://www.reuters.com/article/2015/11/12/us-alibaba-singles-day-idUSKCN0SZ34J20151112#g18I6tqKXwWggva6.97

Cohn & Wolf. (2014). The age of authenticity. cohn & wolf.

Deborah, J. G., & Noorman, M. (2014). Recommendations for Future Development of Artificial Agents. IEEE Technology and Society Magazine, 22–28.

Duggan, M. (2015, 08 19). Mobile Messaging and Social Media 2015. Retrieved 11 18, 2015, from Pew Reserach Center: http://www.pewinternet.org/2015/08/19/mobile-messaging-and-social-media-2015/

Duggan, M. (2015, 8 19). Mobile Messaging and Social Media 2015. Retrieved 11 18, 2015, from Pew Research Center: https://getpocket.com/redirect?url=http%3A%2F%2Fwww.pewinternet.org%2F2015%2F08%2F19%2Fmobile-messaging-and-social-media-2015%2F

Evans, B. (2015, 3 24). Messaging and mobile platforms. Retrieved 11 18, 2015, from Benedict Evans: http://ben-evans.com/benedictevans/2015/3/24/the-state-of-messaging

Fenn, J., & Raskino, M. (2008). Mastering the Hype Cycle: How to Choose the Right Innovation at the Right Time.

Ferrara, E., Varol, O., Davis, C., Menczer, F., & Flammini, A. (2015, 06 2015). The Rise of Social Bots. Retrieved 11 18, 2015, from Cornell University Library: http://arxiv.org/abs/1407.5225

Gartner Inc. (2015, 10 14). Gartner. Retrieved 11 18, 2015, from Five Key Trends in Gartner’s 2015 Digital Marketing Hype Cycle: http://www.gartner.com/smarterwithgartner/five-key-trends-in-gartners-2015-digital-marketing-hype-cycle/

Gulikers, J. T., Bastiaens, T. J., & Kirschner, P. A. (2004). Assessment, A Five-Dimensional Framework for Authentic. ETR&D, 52(3), 67–86.

Haro, J. (2015, 5 27). B2C messaging on WhatsApp: A mobile marketer’s dream. Retrieved 11 18, 2015, from VentureBeat: http://venturebeat.com/2015/05/27/b2c-messaging-on-whatsapp-a-mobile-marketers-dream/

Kolar, T., & Zabkar, V. (2010). A consumer-based model of authenticity: An oxymoron or the foundation of cultural heritage marketing? Tourism Management, 21, 652–664.

Kornienko, S., Kornienko, O., & Levi, P. (2005). Collective AI: context awareness via communication. IJCAI, 5, 1464–1470.

Lee, K., Eoff, B. D., & Caverlee, J. (2011). Seven Months with the Devils: A Long-Term Study of Content Polluters on Twitter. Proceedings of the 5th International AAAI Conference on Weblogs and Social Media, (pp. 185–192).

Lei, H. I. (2005). Context Awareness: a Practitioner’s Perspective. Proceedings of the 2005 International Workshop on Ubiquitous Data Management (UDM’05) (pp. 43–52). Tokyo: IEEE Computer Society.

Leigh, T. W., Peters, C., & Shelton, J. (2006). The Consumer Quest for Authenticity: The Multiplicity of Meanings Within the MG Subculture of Consumption. Journal of the Academy of Marketing Science, 34(4), 481–493.

Lohr, S. (2015, 11 4). Digital Transformation Going Mainstream in 2016, IDC Predicts. Retrieved 11 18, 2015, from The New York Times: http://bits.blogs.nytimes.com/2015/11/04/in-2016-digital-transformation-goes-mainstream-idc-predicts/?_r=0

Mayring, P. (2014). Qualitative content analysis: theoretical foundation, basic procedures and software solution. Klagenfurt.

Olanoff, D., & Constine, J. (2015, 8 26). Facebook Is Adding A Personal Assistant Called “M” To Your Messenger App. Retrieved 11 18, 2015, from Techcrunch: http://techcrunch.com/2015/08/26/facebook-is-adding-a-personal-assistant-called-m-to-your-messenger-app/#.iido8x:dim9

Oskouei, R. J., Varzeghani, H. N., & Samadyar, Z. (2014). Intelligent Agents: A Comprehensive Survey. International Journal of Electronics Communication and Computer Engineering, 5(4), 790–798.

Persson, P., Laaksolahti, J., & Lönnqvist, P. (2001). Understanding Socially Intelligent Agents — A Multilayered Phenomenon. IEEE Transactions on Systems, Man, and Cybernetics — Part A: Systems and Humans, 31(5), 349–360.

Pew Research Center. (2014, February). Couples, the Internet, and Social Media. Retrieved November 2014, from http://pewinternet.org/Reports/2014/Couples-and-the-Internet.aspx

Pine, J., & Gilmore, J. H. (2008). Keep it real. Marketing Management, 18–24.

Prensky, M. (2009, 10). Digital Natives Digital Immigrants. On the Horizon, 9(5).

Rivenbark, L. (2006, May). Social Intelligence. HR Magazine, 141–142.

Sheth, B. (2015, 09 21). Forget Apps! Message Bots Are The Future of Customer Engagement! Retrieved 11 18, 2015, from Techstory: http://techstory.in/message-bots/

Sheth, B. (2015, 9 29). TechCrunch. Retrieved 11 18, 2015, from Forget Apps, Now The Bots Take Over: http://techcrunch.com/2015/09/29/forget-apps-now-the-bots-take-over/

Turing, A. M. (1950). Computing machinery and intelligence. Mind 49, 49(236), 433–460.

Wang, S. (2015, 08 13). The New York Times built a Slack bot to help decide which stories to post to social media. Retrieved 11 18, 2015, from Nieman Foundation at Harvard: http://www.niemanlab.org/2015/08/the-new-york-times-built-a-slack-bot-to-help-decide-which-stories-to-post-to-social-media/

Yücel, R., & Dagelen, O. (2010). Globalization of Markets, Marketing Ethics and Social Responsibility, Globalization — Today, Tomorrow, (InTech ed.). (K. Deng, Ed.) Sciyo.